Hacker News new | past | comments | ask | show | jobs | submit | Scarblac's comments login

61.5% sounds like a lot, but how much electricity is needed to replace everything that fossil fuel is now used for? I think it needs to increase to 400% or so to do that.

We don't need to replace primary energy with electric energy in a 1:1 ratio.

An ICE car is 20-30% efficient, an EV is 90% efficient. Generally we are looking at a grid expansion, but it is not massive.

See this amazing chart on rejected vs. useful energy:

https://flowcharts.llnl.gov/


> an EV is 90% efficient

An EV is 85-90% efficient with power coming out of the batteries. However, the charging efficiency of the batteries is at about 80%. So the overall efficiency is 80% of 90%, or 72%.


For the sake of replacement, we should not include the charging efficiency, because the context is that of "how kuch energy the car can store" rather than "where it comes from"

It's "how much is needed to keep doing the same thing", I don't see why charging would be excluded.

It is not about how much it can store. You cannot wish away the charging losses. There are no losses to pumping gas into a tank. Nor can you wish away the losses in the charger itself. Converting AC to DC is hardly 100% efficient.

In that case, you should also not wish away refining losses, where petroleum is turned into gasoline and other fuels.

If you want to get an accurate accounting of CO2 emissions, sure.

Somewhere between 2-3x the electricity consumed right now (which is way down from the peak several years ago as efficiency increases, e.g. LEDs instead of light bulbs etc.). It would be crazy expensive to build nuclear reactors to produce so much electricity, with renewables it is still quite a task, but a much more manageable one.

in fact for germany building a similar renewable output will be much more expensive. Don't forget about transmission and balancing. Needless to say Germany doesn't have an actual plan to ditch gas, even h2-ready plants will either use a mix with gas(most probably) or pure h2(if it'll even be deployed) will still have huge NOx emissions

Sounds good, we've gone only in the opposite direction for far too long.

There's also chores, talking with people, board games, going for a walk, a bath, sex, exercise, just doing nothing in particular for a bit, etc. The choice isnt only screens or books.

Doing the dishes is a nice balance of meditative, boring, and feeling accomplished. A good nightly chore.

Exercise can backfire in terms of sleep. A bit earlier in the day and it's great. But not right before bed, at least not in my experience.


So how many 1 bits are in strawberry?

Thats about the same kind of question, given that LLMs see tokens instead of characters, and we see characters instead of bits.


Do you sincerely think the folks training these models haven't tried alternative tokenizations?

Of course they have, I assume counting letters in words wasn't a priority for them.

> Meanwhile, humans in general are pretty terrible at exact, instantaneous arithmetic. But we aren’t claiming that computers are intelligent because they’re great at it.

"The question of whether a computer can think is no more interesting than the question of whether a submarine can swim." - Edsger Dijkstra


The smallest known insects are also parasitic wasps: https://en.m.wikipedia.org/wiki/Dicopomorpha_echmepterygis

They're wild. They're born inside the eggs of barklice, and mate there. Males are ~186 micrometer, smaller than some single celled organisms like amoeba!


They decided to redistribute their modified version under the terms of the GPLv3, as they were allowed to do.

The recipients thus receive software licensed under the GPLv3.

I dont see why people would have any GPLv2 rights for this version.


> I dont see why people would have any GPLv2 rights for this version.

Because the code is licensed under GPL v2. Not or any later versions. In the original distribution there was a filed called LICENSE.txt which is just the literal GPL v2.0 text, which includes the text of the license as well as the explanatory text and instructions on how to apply it. There is a README.txt which includes the line "Licensed under the GNU General Public License 2.0." And then every source code file says:

     // This source is available for distribution and/or modification
     // only under the terms of the DOOM Source Code License as
     // published by id Software. All rights reserved.
So you would have the choice of GPLv2 or possibly this: https://doomwiki.org/wiki/Doom_Source_License


Wondering what people's opinion on the _Domain Driven Design_ book is.


I'll be honest. I hated it. Any book that contains the text "obviously ..." loses my trust. I can't respect an author who's too lazy to make his point.

He used it quite a few times in the beginning of the book.

"Obviously" is used for two reasons. Derision, or to escape criticism.

I'm obviously right.

Now the real critique. The book assumes experts are on hand to ask, and that terms don't change. I worked at a place where what we called user changed 3 times in 4 years. Runners/workers/producers. You would know when a bit of code was written based on the names for users in the api.

It also assumes that all experts agree. I worked on an education app where teachers in schools were the experts. Different schools in the same city had different terms for the same thing. So which expert do you trust.

Is it a trunk or a boot? Is it an elevator or a lift? And that's if you're writing an app for english only speaking experts.

Finally, we have to understand the mind of the dev, to know how they organised their walls. The dev that left the company 5 years ago.


I'm eager to hear other opinions on Eric's DDD book too :-D I still don't understand it!


Waist of time.


It's a nice system but far too complicated for the world we live in, where most tech companies, especially banks, are filled with imported labor or outsourced globally.

DDD requires a certain discipline and dedication I just haven't seen in my career across many companies.

Good luck getting more than a handful of people to not only digest the big books but also structure all their code and company around practicing it.


How many of the US brightest go work in aerospace engineering these days?

Edit: and to those that do, SpaceX probably looks a lot more attractive.


Probably not a lot when in the same city you have two huge tech companies. You would be crazy to go to Boeing


At first, that new term is not just the "updated term" that everyone knows is now considered less offensive.

At first it's partly activist, and when you adopt it it means that many people won't understand you, think you're criticizing their use of the previous term, or think you're probably activist on other topics too, when you think you're using a normal, neutral expression.

So most people are sceptical and wait with adopting new terms until they're widespread.


Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: